Page 24
How to achieve the impossible
in an Emergency Department
Dr. Phil Munro describes how a programme in Glasgow
to admit or discharge 98 per cent of all Emergency
Department patients within four hours from time
of registration was successful beyond all expectations
When the Scottish Government
Health Department (SGHD) unveiled
the Unscheduled Care Collaborative
Programme it was met with equal
measures of enthusiasm and trepidation. Our target for December 2007
was that 98 per cent of all Emergency
Department patients should be admitted or discharged within four hours
from time of registration. An interim
target of 95 per cent was to be met by
December 2006.
This was an ambitious (and indeed
many felt impossible) target, but we
now had a clear governmental and
managerial mandate to achieve this
and were tasked with addressing it in
our own hospital and across Glasgow.
Those of us working in emergency
medicine saw it as tremendous opportunity to improve the speed and efficiency of patient services. We also
saw it as a potential means of securing additional resources for emergency care in a climate where for
many years large amounts of funding
had been poured into elective care
and waiting list initiatives.
It soon became clear that, at least
in the short term, no major investment was likely other than in IT support and management facilitation of
the programme. Having been interested for some time in the concepts of
“lean” as applied to emergency
health care I felt there we had to initially explore resource – neutral
process redesign.
The first component addressed was
that of information. We had been
closely monitoring our ED waiting
times for several years and noted
their depressingly inexorable decline.
Despite some increase in staffing and
a high quality teaching programme,
we were faced daily with spiralling
problems of “exit block”; a term used
to denote patients on trolleys awaiting
admission to an in-patient ward.
Sequential lines in the sand about
patients spending overnight on trolleys in the ED had been crossed and
we were seeing the resultant rise in
critical incidents and complaints.
We needed to present the information in a clear, easily charted and
unambiguous fashion. Fortunately
the SGHD had been careful to define
“admission” such that we could not
simply call a hallway or storage room
a ward and move trolleys in to these
areas thus “stopping the clock” as
had occurred elsewhere in the UK.
We used three primary measures;
percentage of patients seen, assessed,
treated and discharged from the
emergency department within four
hours of registration; the number who
were not discharged or admitted within four hours (“breachers”) and also
initially patients who breached 12
hours. These were used to performance manage the ED as well as inpatient specialties and clinical service managers by subdividing them
into minor injuries, medical and surgical “flows”. A crucial aspect was
that these had to be seen as “whole
hospital” targets and not simply an
ED problem.
From the outset, we were careful to
produce regular accurate performance
figures and disseminated these very
widely to all areas and specialties.
Initially these were distributed monthly, then weekly, then daily and, at the
time of writing, we are updated twice
daily with a table of the performance
of our nine regional hospitals.
Clinical leaders
We then set about improving all aspects
of patient management, carefully
involving clinical leaders, senior managers and support services. We tried to
incorporate lean principles of reducing
duplication, reducing waste, adding
value and simplifying or abolishing
queues where possible. These included:
Value Added Triage to avoid
triage simply acting as a meet and
Page 25
Emergency | targets
greet or at worst a bottleneck. We
taught nursing staff to order
selected x-rays according to strict
criteria, dispense simple
painkillers and to be able to redirect specific patients back to their
general practitioners.
A demand analysis from arrival
times in the emergency department allowing us to rearrange our
medical staffing to better match
workload. For example we moved
staff from the less busy Thursday
and Friday afternoons to the much
busier Monday and Tuesday
Removing the general medicine
interns from the ED as they did
not make decisions on admission
or discharge and their work was
subsequently duplicated by one or
more levels of more senior staff
before patients were moved on.
Separating minor injuries into a
separate stream and ring fencing
Emergency Nurse Practitioners
(ENP) to see, treat and discharge
these patients. As a result minor
injuries continue to be processed
efficiently even when complex
resuscitation cases were consuming senior medical resources.
Creating a new ED admission
document along with our colleagues in general medicine to act
as an admission note for up to 72
hours. This dramatically cut
duplication of admission paper
work and streamlined the whole
process of emergency admission.
Creating a work board for portering staff to coordinate and prioritise tasks along with the nurse in
charge of the ED.
Combining all patients arriving
with a general medical complaint
(the most common group requiring
admission) into a single queue.
Previously, patients referred by
general practitioners were seen
separately by in-patient teams.
By removing this distinction we
were able to move to patients
being admitted or discharged after
a single complete assessment in
the ED as opposed to two or more
before a decision was made.
Establishing a daily bed meeting
led by our bed managers and
attended by senior medical and
nursing staff from the ED, senior
nursing staff from the main
admitting wards. These involve a
rapid review of the previous day’s
performance, current bed states
including expected discharges,
planned admissions and predicted emergency admissions resulting in a daily plan which is
shared with our clinical services
Allocation of Estimated Date of
Discharge (EDD) for each admitted patient. This can be identified within 24 hours in most cases
and is flexible enough to change if
the patient’s condition alters.
Patients and their relatives are
given written and verbal instructions that discharges will normally
be before 12 noon on the day of
These required no substantial
investment other than a commitment
to change and the enthusiasm of the
medical, nursing and managerial
staff to make these systems sustainable. We have subsequently invested in increased staffing for our ENP
service to ensure consistent manning
and we created a nine bedded
Clinical Decision Unit (CDU) where
highly selected medical patients are
processed in a nurse-led protocol
driven short stay unit (currently
open Monday to Thursday). This
resulted in an approximately 24 hour
reduction in the length of stay for
these conditions and along with the
use of EDD has virtually abolished
boarding medical patients into surgical wards.
It is important to acknowledge our
failed initiatives including a patient
discharge lounge and a trial of rapid
access consultant clinics to provide
an alternative to admission that had
no impact at all on throughput.
Despite this the programme has
been successful beyond all expectations. From August 2005 our average
ED annual attendance rose from
46,000 to 47,887. During this time
our average performance against the
four hour target rose from 84.7 per
cent to 95 per cent and at the time of
writing it is holding steady between
98 per cent – 99 per cent.
We could not simply
call a hallway or
storage room a ward
and move trolleys in
to these areas thus
“stopping the clock”
The clear message I would give to
anyone faced with seemingly
intractable problems of ED waiting
times, exit block and lack of emergency admission capacity is that an
identical situation was considered
impossible to fix in Glasgow. Some
investment in staffing and targeted
interventions will be required but dramatic improvements can be achieved
with creative re-engineering of admission processes led by a committed
clinical and managerial team.
Consultant and Honorary Clinical Senior
Emergency Medicine
Southern General Hospital
[email protected]